Needle exchanges are innovative but controversial interventions designed to reduce the spread of the Human Immunodeficiency Virus (HIV) among injection drug users (IDUs). Proponents argue that they are the most direct and powerful measure we have for achieving that goal, while critics say that such programs condone and encourage drug use. Objective evaluation of needle exchange programs is crucial to a resolution of this debate, and to provide guidance to public health policy. We propose a longitudinal cohort study of the impact of Seattle - King County's Needle Exchange, which has been in operation since April of 1989, and which has exchanged over 608,000 syringes. IDUs who have used the needle exchange (exposed cohort) will be compared to IDUs who have not used the needle exchange (unexposed cohort) to observe changes in the practice of high risk sexual and drug injecting behaviors, the acquisition of sexually and parenterally acquired infections, changes in frequency of drug injection, entrance into drug treatment, and changes in knowledge of HIV transmission and its prevention. Steps will be taken in the design and analysis to minimize, assess, and control for the potential bias introduced by self-selection to exchange patronage, and by possible differential follow-up among exchangers and non-exchangers. Logistic regression modeling will enable control for potential confounders.